CN104550253A - Pre-warning system and pre-warning method for inclination of temper mill - Google Patents

Abstract

The invention relates to a cold-rolled sheet production system, in particular to a temper mill. A pre-warning system for inclination of the temper mill is characterized in that whether the temper mill may be in inclination or not can be subjected to computation processing by a central processing machine; a result can be sent to the temper mill if an abnormal situation appears, and whistling warning can be carried out while a lamp flashes. The pre-warning system is formed by two parts, wherein one part is provided with the central processing machine which is responsible for analyzing and forecasting the inclination of the temper mill, and the other part is provided with a warning lamp and a buzzer which are distributed in a roller group of the temper mill; flashing and whistling can be carried out for reminding a worker to operate when an inclination alarm of the temper mill is received. According to the pre-warning system disclosed by the invention, the problem of the inclination of the temper mill can be effectively solved; the quality of strip steel products is ensured, and further generation of strip steel defects is reduced by increasing the production benefit through the effective improvement.

Description

A kind of planisher inclination early warning system and method for early warning thereof

Technical field

The present invention relates to flat cold-rolled sheet production system, particularly relate to planisher.

Background technology

In flat cold-rolled sheet production technology, smooth is first operation after the annealing of band steel, the yield point elongation after major function has cancellation band steel to anneal, provides necessary mechanical property, improves template and controls the brightness of belt steel surface and extra coarse degree etc.The whether normal of planisher work has very important impact to strip quality, and system carries out early warning to the roll-force of planisher and inclination fault for this reason, thus at utmost reduces the issuable loss of unit.

Temper mill roller inclination Producing reason has several, more commonly the difference in thickness of two continuous supplied materials is excessive, when a coil volume is after planisher, planisher two knobbling rolls spacing does not change, and next coil volume thickness exceedes last coil when licking many, temper mill roller pressure at two ends deviation may be caused, when rolling force deviation is excessive, then can produces coil of strip limit wave, template is not inconsistent or the problems such as roll surface of wearing and tearing.

How effectively to judge that temper mill roller inclination fault becomes the emphasis of planisher Processes and apparatus assurance.Consider from temper mill roller roll surface flatness angle in the past, think and cause roll surface unbalance stress because roll flatness is not inconsistent and causes rolling force deviation.But usually also there will be inclination fault after roll change, can not the economic loss brought of timely less fault.

Summary of the invention

The present invention is intended to address the aforementioned drawbacks, and provides a kind of planisher inclination early warning system and method for early warning thereof.The present invention effectively can solve the problem that planisher tilts.Present invention ensures that belt steel product quality, improve productivity effect by effectively improving, decreasing the further generation of steel defect.

For solving the problem, a kind of planisher inclination early warning system, it comprises the central processing unit of responsible planisher pitch analysis and prediction, the innage position of roll driving side roll-force and fore side roll-force and roll driving side and fore side monitored by this central processing unit simultaneously, it also comprises alarm lamp and the whistling device of central processing unit control, alarm lamp and whistling device have two groups, are located at the both sides between No. 1 and No. 2 planishers separately and arrange near band steel walking position.

The method for early warning of described planisher inclination early warning system, it comprises:

Step one, CPU, to planisher pitch analysis, comprises and extracts the generation that following production rule judges roller declination fault, and this rule, for reducing wrong report, have employed continuous multi-point sampling decision principle;

1) data acquisition and pretreatment

The rolling force F WS of Real-time Collection roll driving side and fore side ^*, FDS ^*and the innage position of roll driving side and fore side is YWS ^*, YDS ^*.

Eccentric for removing roller, the impact of bear vibration etc., the following moving average filter process of above data acquisition:

$F_{\mathrm{ws}}\left(k\right)=\frac{\underset{j=1}{\overset{N-1}{\mathrm{\Σ}}}{F}_{\mathrm{ws}}^{*}(k-j)}{N},F_{\mathrm{ds}}\left(k\right)=\frac{\underset{j=0}{\overset{N-1}{\mathrm{\Σ}}}{F}_{\mathrm{ds}}^{*}(k-j)}{N}$

$Y_{\mathrm{ws}}\left(k\right)=\frac{\underset{j=1}{\overset{N-1}{\mathrm{\Σ}}}{Y}_{\mathrm{ws}}^{*}(k-j)}{N},Y_{\mathrm{ds}}\left(k\right)=\frac{\underset{j=0}{\overset{N-1}{\mathrm{\Σ}}}{Y}_{\mathrm{ds}}^{*}(k-j)}{N}$

Wherein, N is moving average step-length, elects 10 as;

2) roller declination failure prediction is according to: roller declination rule, judges roller declination state;

IF |FWS(k)-FDS(k)|>X _aAND |YWS(k)-YDS(k)|>X _bAND|FWS(k+1)-FDS(k+1)|>X _aAND |YWS(k+1)-YDS(k+1)|>X _bAND|FWS(k+m)-FDS(k+m)|>X _aAND |YWS(k+m)-YDS(k+m)|>X _b

Then reports to the police

ELSE is normal

Wherein, FWS, FDS represent roll driving side roll-force and fore side roll-force respectively; YWS, YDS represent the innage position of roll driving side and fore side respectively; M is continuous sampling number;

When temper mill roller keeps stablizing slave mode by analysis, roll-force and innage positioning index are normal distribution, so setup control circle, making the idea in control chart have 99.73% drops in control limit, in center line both sides random distribution, that is, control limit is set to center line ± 3 times standard deviation, draws according to above method and control boundary: X _a=25t, X _b=0.7mm, the threshold value that historical failure data that Here it is is determined by 95% confidence limit;

Step 2, alarming result is sent to alarm lamp and whistling device by network by CPU, and alarm lamp and whistling device start to report to the police.

Present invention optimizes the account form of early warning value, the generation of this algorithm will greatly reduce the human factor of early warning optimization, reduce the possibility of broken belt, sideslip, remind field technician to monitor band steel, check, analyze in time, avoid unnecessary loss occurrence; Secondary machine directly intercepts tension signal in PLC, directly judges whether to reach Alert Standard, greatly accelerates early warning speed, meets the demand of high speed rolling mill production line, to ensureing carrying out smoothly and the quality important in inhibiting of finished strip of the operation of rolling.

Accompanying drawing illustrates:

Below in conjunction with accompanying drawing, the present invention is further illustrated:

Fig. 1 is early warning system schematic diagram of the present invention;

Fig. 2 is roll-force normal distribution curve schematic diagram;

Fig. 3 is innage position normal distribution curve schematic diagram.

Detailed description of the invention

Refer to Fig. 1, a kind of planisher inclination early warning system, it comprises the central processing unit 1 of responsible planisher pitch analysis and prediction, the innage position of roll driving side roll-force and fore side roll-force and roll driving side and fore side monitored by this central processing unit simultaneously, it also comprises alarm lamp 2 and the whistling device 3 of central processing unit control, alarm lamp and whistling device have two groups, are located at the both sides between No. 1 and No. 2 planishers separately and arrange near band steel walking position.

The method for early warning of described planisher inclination early warning system, it comprises:

Step one, CPU, to planisher pitch analysis, comprises and extracts the generation that following production rule judges roller declination fault, and this rule, for reducing wrong report, have employed continuous multi-point sampling decision principle;

1) data acquisition and pretreatment

The rolling force F WS of Real-time Collection roll driving side and fore side ^*, FDS ^*and the innage position of roll driving side and fore side is YWS ^*, YDS ^*,

Eccentric for removing roller, the impact of bear vibration etc., the following moving average filter process of above data acquisition:

$F_{\mathrm{ws}}\left(k\right)=\frac{\underset{j=1}{\overset{N-1}{\mathrm{\Σ}}}{F}_{\mathrm{ws}}^{*}(k-j)}{N},F_{\mathrm{ds}}\left(k\right)=\frac{\underset{j=0}{\overset{N-1}{\mathrm{\Σ}}}{F}_{\mathrm{ds}}^{*}(k-j)}{N}$

$Y_{\mathrm{ws}}\left(k\right)=\frac{\underset{j=1}{\overset{N-1}{\mathrm{\Σ}}}{Y}_{\mathrm{ws}}^{*}(k-j)}{N},Y_{\mathrm{ds}}\left(k\right)=\frac{\underset{j=0}{\overset{N-1}{\mathrm{\Σ}}}{Y}_{\mathrm{ds}}^{*}(k-j)}{N}$

Wherein, N is moving average step-length, elects 10 as;

2) roller declination failure prediction is according to: roller declination rule, judges roller declination state;

IF |FWS(k)-FDS(k)|>X _aAND |YWS(k)-YDS(k)|>X _bAND|FWS(k+1)-FDS(k+1)|>X _aAND |YWS(k+1)-YDS(k+1)|>X _bAND|FWS(k+m)-FDS(k+m)|>X _aAND |YWS(k+m)-YDS(k+m)|>X _b

Then reports to the police

ELSE is normal

Wherein, FWS, FDS represent roll driving side roll-force and fore side roll-force respectively; YWS, YDS represent the innage position of roll driving side and fore side respectively; M is continuous sampling number;

When temper mill roller keeps stablizing slave mode by analysis, roll-force and innage positioning index are normal distribution (see Fig. 2 and Fig. 3), so setup control circle, making the idea in Fig. 2 and Fig. 3 have 99.73% drops in control limit, in center line both sides random distribution, that is, control limit is set to center line ± 3 times standard deviation, draws according to above method and control boundary: X _a=25t, X _b=0.7mm, the threshold value that historical failure data that Here it is is determined by 95% confidence limit;

Step 2, alarming result is sent to alarm lamp and whistling device by network by CPU, and alarm lamp and whistling device start to report to the police.

Present invention optimizes the account form of early warning value, the generation of this algorithm will greatly reduce the human factor of early warning optimization, reduce the possibility of broken belt, sideslip, remind field technician to monitor band steel, check, analyze in time, avoid unnecessary loss occurrence; Secondary machine directly intercepts tension signal in PLC, directly judges whether to reach Alert Standard, greatly accelerates early warning speed, meets the demand of high speed rolling mill production line, to ensureing carrying out smoothly and the quality important in inhibiting of finished strip of the operation of rolling.

Claims (2)

1. a planisher inclination early warning system, it comprises the central processing unit of responsible planisher pitch analysis and prediction, the innage position of roll driving side roll-force and fore side roll-force and roll driving side and fore side monitored by this central processing unit simultaneously, it is characterized in that, it also comprises alarm lamp and the whistling device of central processing unit control, alarm lamp and whistling device have two groups, are located at the both sides between No. 1 and No. 2 planishers separately and arrange near band steel walking position.

2. a method for early warning for planisher inclination early warning system according to claim 1, it comprises:

Step one, CPU, to planisher pitch analysis, comprises and extracts the generation that following production rule judges roller declination fault, and this rule, for reducing wrong report, have employed continuous multi-point sampling decision principle;

1) data acquisition and pretreatment

The rolling force F WS of Real-time Collection roll driving side and fore side ^*, FDS ^*and the innage position of roll driving side and fore side is YWS ^*, YDS ^*;

Eccentric for removing roller, the impact of bear vibration etc., the following moving average filter process of above data acquisition:

$F_{\mathrm{ws}}\left(k\right)=\frac{\underset{j=1}{\overset{N-1}{\mathrm{\Σ}}}{F}_{\mathrm{ws}}^{*}(k-j)}{N},F_{\mathrm{ds}}\left(k\right)=\frac{\underset{j=0}{\overset{N-1}{\mathrm{\Σ}}}{F}_{\mathrm{ds}}^{*}(k-j)}{N}$

$Y_{\mathrm{ws}}\left(k\right)=\frac{\underset{j=1}{\overset{N-1}{\mathrm{\Σ}}}{Y}_{\mathrm{ws}}^{*}(k-j)}{N},Y_{\mathrm{ds}}\left(k\right)=\frac{\underset{j=0}{\overset{N-1}{\mathrm{\Σ}}}{Y}_{\mathrm{ds}}^{*}(k-j)}{N}$

Wherein, N is moving average step-length, elects 10 as;

2) roller declination failure prediction is according to: roller declination rule, judges roller declination state;

IF |FWS(k)-FDS(k)|>Xa AND |YWS(k)-YDS(k)|>Xb AND|FWS(k+1)-FDS(k+1)|>Xa AND |YWS(k+1)-YDS(k+1)|>Xb…AND|FWS(k+m)-FDS(k+m)|>Xa AND |YWS(k+m)-YDS(k+m)|>Xb

Then reports to the police

ELSE is normal

Wherein, FWS, FDS represent roll driving side roll-force and fore side roll-force respectively; YWS, YDS represent the innage position of roll driving side and fore side respectively; M is continuous sampling number;

When temper mill roller keeps stablizing slave mode by analysis, roll-force and innage positioning index are normal distribution, so setup control circle, making the idea in control chart have 99.73% drops in control limit, in center line both sides random distribution, that is, control limit is set to center line ± 3 times standard deviation, draws according to above method and control boundary: X _a=25t, X _b=0.7mm, the threshold value that historical failure data that Here it is is determined by 95% confidence limit;

Step 2, alarming result is sent to alarm lamp and whistling device by network by CPU, and alarm lamp and whistling device start to report to the police.